What is Llama 2?

Software needs ways to work with ordinary human language: writing, summarizing, translating and analysing text. Large language models are one approach: neural networks trained on a vast amount of text, which can typically generate, summarize, translate and analyze text in many contexts. Llama 2 is a family of such models from Meta, described in its paper as pretrained and fine-tuned language models ranging from 7 billion to 70 billion parameters (parameters are the learned numbers that make up a model). The fine-tuned versions (further trained for a narrower purpose), called Llama 2-Chat, are optimized for dialogue, meaning back-and-forth conversation with a user. Meta's license page gives a version release date of July 18, 2023.

Written from primary sources and reference pages, and checked against them by an independent AI fact-checker. Drafted with AI assistance.